Big Debt Crises - PDFDrive.comRay Dalio
Table of Contents
A Template for Understanding Big Debt Crises Acknowledgements Introduction PART 1: The Archetypal Big Debt Cycle How I Think about Credit and Debt The Template for the Archetypal Long-Term/Big Debt Cycle Our Examination of the Cycle The Phases of the Classic Deflationary Debt Cycle The Early Part of the Cycle The Bubble The Top The “Depression” The “Beautiful Deleveraging” “Pushing on a String” Normalization Inflationary Depressions and Currency Crises The Phases of the Classic Inflationary Debt Cycle The Early Part of the Cycle The Bubble The Top and Currency Defense The Depression (Often When the Currency Is Let Go) Normalization The Spiral from a More Transitory Inflationary Depression to Hyperinflation War Economies In Summary PART 2: Detailed Case Studies German Debt Crisis and Hyperinflation (1918-1924) US Debt Crisis and Adjustment (1928–1937) US Debt Crisis and Adjustment (2007–2011) PART 3: Compendium of 48 Case Studies Glossary of Key Economic Terms Primarily Domestic Currency Debt Crises Non-Domestic Currency Debt Crises Appendix: Macroprudential Policies
Acknowledgements
I cannot adequately thank the many people at Bridgewater who have shared, and continue to share, my mission to understand the markets and to test that understanding in the real world. I celebrate the meaningful work and meaningful relationships that we have had and that have led to the understandings and principles that have enriched us in the most profound ways.
The list of great partners includes Bob Prince, Greg Jensen and Dan Bernstein, who have worked with me over decades, my current investment research team (especially Steven Kryger, Gardner Davis, Bill Longfield, Anser Kazi, Danny Newman, Michael Savarese, and Elena Gonzalez Malloy), members of my past research teams (especially Brian Gold, Claude Amadeo, Bob Elliott, Mark Dinner, Brandon Rowley, and Jason Rogers), and many others who have worked with me on research over the years. I’m also indebted to the many other leaders in Bridgewater research, including Jason Rotenberg, Noah Yechiely, Larry Cofsky, Ramsen Betfarhad, Karen Karniol-Tambour, Kevin Brennan, Kerry Reilly, Jacob Kline, Avraam Sidiropoulos, Amit Srivastava, and our treasured former colleague Bruce Steinberg, who we tragically lost last year.
Introduction
I am writing this on the tenth anniversary of the 2008 financial crisis in order to offer the perspective of an investor who navigated that crisis well because I had developed a template for understanding how all debt crises work. I am sharing that template here in the hope of reducing the likelihood of future debt crises and helping them be better managed.
As an investor, my perspective is different from that of most economists and policy makers because I bet on economic changes via the markets that reflect them, which forces me to focus on the relative values and flows that drive the movements of capital. Those, in turn, drive these cycles. In the process of trying to navigate them, I’ve found there is nothing like the pain of being wrong or the pleasure of being right as a global macro investor to provide the practical lessons about economics that are unavailable in textbooks.
After repeatedly being bit by events I never encountered before, I was driven to go beyond my own personal experiences to examine all the big economic and market movements in history, and to do that in a way that would make them virtual experiences—i.e., so that they would show up to me as though I was experiencing them in real time. That way I would have to place my market bets as if I only knew what happened up until that moment. I did that by studying historical cases chronologically and in great detail, experiencing them day by day and month by month. This gave me a much broader and deeper perspective than if I had limited my perspective to my own direct experiences. Through my own experience, I went through the erosion and eventual breakdown of the global monetary system (“Bretton Woods”) in 1966–1971, the inflation bubble of the 1970s and its bursting in 1978–82, the Latin American inflationary
depression of the 1980s, the Japanese bubble of the late 1980s and its bursting in 1988–1991, the global debt bubbles that led to the “tech bubble” bursting in 2000, and the Great Deleveraging of 2008. And through studying history, I experienced the collapse of the Roman Empire in the fifth century, the United States debt restructuring in 1789, Germany’s Weimar Republic in the 1920s, the global Great Depression and war that engulfed many countries in the 1930–45 period, and many other crises.
My curiosity and need to know how these things work in order to survive them in the future drove me to try to understand the cause-effect relationships behind them. I found that by examining many cases of each type of economic phenomenon (e.g., business cycles, deleveragings) and plotting the averages of each, I could better visualize and examine the cause-effect relationships of each type. That led me to create templates or archetypal models of each type—e.g., the archetypal business cycle, the archetypal big debt cycle, the archetypal deflationary deleveraging, the archetypal inflationary deleveraging, etc. Then, by noting the differences of each case within a type (e.g., each business cycle in relation to the archetypal business cycle), I could see what caused the differences. By stitching these templates together, I gained a simplified yet deep understanding of all these cases. Rather than seeing lots of individual things happening, I saw fewer things happening over and over again, like an experienced doctor who sees each case of a certain type of disease unfolding as “another one of those.”
I did the research and developed this template with the help of many great partners at Bridgewater Associates. This template allowed us to prepare better for storms that had never happened to us before, just as one who studies 100- year floods or plagues can more easily see them coming and be better prepared. We used our understanding to build computer decision-making systems that laid out in detail exactly how we’d react to virtually every possible occurrence. This
approach helped us enormously. For example, eight years before the financial crisis of 2008, we built a “depression gauge” that was programmed to respond to the developments of 2007–2008, which had not occurred since 1929–32. This allowed us to do very well when most everyone else did badly.
While I won’t get into Bridgewater’s detailed decision making systems, in this study I will share the following: 1) my template for the “Archetypal Big Debt Cycle;” 2) “Three Iconic Case Studies” examined in detail (the US in 2007– 2011, which includes the “Great Recession;” the US in 1928–1937, which covers a deflationary depression; and Germany in 1918–1924, which examines an inflationary depression), and 3) a “Compendium of 48 Case Studies,” which includes most of the big debt crises that happened over the last 100 years.* I guarantee that if you take the trouble to understand each of these three perspectives, you will see big debt crises very differently than you did before.
To me, watching the economy and markets, or just about anything else, on a day-to-day basis is like being in an evolving snowstorm with millions of bits and pieces of information coming at me that I have to synthesize and react to well. To see what I mean by being in the blizzard versus seeing what’s happening in more synthesized ways, compare what’s conveyed in Part 1 (the most synthesized/template version) with Part 2 (the most granular version), and Part 3 (the version that shows the 48 cases in chart form). If you do that, you will note how all of these cases transpire in essentially the same way as described in the archetypal case while also noting their differences, which will prompt you to ponder why these differences exist and how to explain them, which will advance your understanding. That way, when the next crisis comes along, you will be better prepared to deal with it.
To be clear, I appreciate that different people have different perspectives, that mine is just one, and that by putting our perspectives out there for debate we can all advance our understandings. I am sharing this study to do just that.
* There is also a glossary of economic terms at the start of Part 3, and for a general overview of many of the concepts contained in this study, I recommend my 30-minute animated video, “How the Economic Machine Works,” which can be accessed at www.economicprinciples.org.
PART 1: The Archetypal Big Debt Cycle
How I Think about Credit and Debt
The Template for the Archetypal Long-Term/Big Debt Cycle
Our Examination of the Cycle
The Phases of the Classic Deflationary Debt Cycle The Early Part of the Cycle • The Bubble • The Top • The “Depression” • The “Beautiful Deleveraging” • “Pushing on a String” • Normalization
Inflationary Depressions and Currency Crises
The Phases of the Classic Inflationary Debt Cycle
The Early Part of the Cycle • The Bubble • The Top and Currency Defense • The Depression (Often When the Currency Is Let Go) • Normalization
The Spiral from a More Transitory Inflationary Depression to Hyperinflation
War Economies
In Summary
How I Think about Credit and Debt
Since we are going to use the terms “credit” and “debt” a lot, I’d like to start with what they are and how they work.
Credit is the giving of buying power. This buying power is granted in exchange for a promise to pay it back, which is debt. Clearly, giving the ability to make purchases by providing credit is, in and of itself, a good thing, and not providing the power to buy and do good things can be a bad thing. For example, if there is very little credit provided for development, then there is very little development, which is a bad thing. The problem with debt arises when there is an inability to pay it back. Said differently, the question of whether rapid credit/debt growth is a good or bad thing hinges on what that credit produces and how the debt is repaid (i.e., how the debt is serviced).
Almost by definition, financially responsible people don’t like having much debt. I understand that perspective well because I share it. For my whole life, even when I didn’t have any money, I strongly preferred saving to borrowing, because I felt that the upsides of debt weren’t worth its downsides, which is a perspective I presume I got from my dad. I identify with people who believe that taking on a little debt is better than taking on a lot. But over time I learned that that’s not necessarily true, especially for society as a whole (as distinct from individuals), because those who make policy for society have controls that individuals don’t. From my experiences and my research, I have learned that too little credit/debt growth can create as bad or worse economic problems as having too much, with the costs coming in the form of foregone opportunities.
1
Generally speaking, because credit creates both spending power and debt, whether or not more credit is desirable depends on whether the borrowed money is used productively enough to generate sufficient income to service the debt. If that occurs, the resources will have been well allocated and both the lender and the borrower will benefit economically. If that doesn’t occur, the borrowers and the lenders won’t be satisfied and there’s a good chance that the resources were poorly allocated.
In assessing this for society as a whole, one should consider the secondary/indirect economics as well as the more primary/direct economics. For example, sometimes not enough money/credit is provided for such obviously cost-effective things as educating our children well (which would make them more productive, while reducing crime and the costs of incarceration), or replacing inefficient infrastructure, because of a fiscal conservativism that insists that borrowing to do such things is bad for society, which is not true.
I want to be clear that credit/debt that produces enough economic benefit to pay for itself is a good thing. But sometimes the trade-offs are harder to see. If lending standards are so tight that they require a near certainty of being paid back, that may lead to fewer debt problems but too little development. If the lending standards are looser, that could lead to more development but could also create serious debt problems down the road that erase the benefits. Let’s look at this and a few other common questions about debt and debt cycles.
How Costly Is Bad Debt Relative to Not Having the Spending that the Debt Is Financing? Suppose that you, as a policy maker, choose to build a subway system that costs $1 billion. You finance it with debt that you expect to be paid back from revenue, but the economics turn out to be so much worse than you expected that only half of the expected revenues come in. The debt has to be written down by 50 percent. Does that mean you shouldn’t have built the subway?
Rephrased, the question is whether the subway system is worth $500 million more than what was initially budgeted, or, on an annual basis, whether it is worth about 2 percent more per year than budgeted, supposing the subway system has a 25-year lifespan. Looked at this way, you may well assess that having the subway system at that cost is a lot better than not having the subway system.
To give you an idea of what that might mean for an economy as a whole, really bad debt losses have been when roughly 40 percent of a loan’s value couldn’t be paid back. Those bad loans amount to about 20 percent of all the outstanding loans, so the losses are equal to about 8 percent of total debt. That total debt, in turn, is equal to about 200 percent of income (e.g., GDP), so the shortfall is roughly equal to 16 percent of GDP. If that cost is “socialized” (i.e., borne by the society as a whole via fiscal and/or monetary policies) and spread over 15 years, it would amount to about 1 percent per year, which is tolerable. Of course, if not spread out, the costs would be intolerable. For that reason, I am asserting that the downside risks of having a significant amount of debt depends a lot on the willingness and the ability of policy makers to spread out the losses arising from bad debts. I have seen this in all the cases I have lived through and studied. Whether policy makers can do this depends on two factors: 1) whether the debt is denominated in the currency that they control and 2) whether they have influence over how creditors and debtors behave with each other.
Are Debt Crises Inevitable? Throughout history only a few well-disciplined countries have avoided debt crises. That’s because lending is never done perfectly and is often done badly due to how the cycle affects people’s psychology to produce bubbles and busts. While policy makers generally try to get it right, more often than not they err on the side of being too loose with credit because the near-term rewards (faster
growth) seem to justify it. It is also politically easier to allow easy credit (e.g., by providing guarantees, easing monetary policies) than to have tight credit. That is the main reason we see big debt cycles.
Why Do Debt Crises Come in Cycles? I find that whenever I start talking about cycles, particularly big, long-term cycles, people’s eyebrows go up; the reactions I elicit are similar to those I’d expect if I were talking about astrology. For that reason, I want to emphasize that I am talking about nothing more than logically-driven series of events that recur in patterns. In a market-based economy, expansions and contractions in credit drive economic cycles, which occur for perfectly logical reasons. Though the patterns are similar, the sequences are neither pre-destined to repeat in exactly the same ways nor to take exactly the same amount of time.
To put these complicated matters into very simple terms, you create a cycle virtually anytime you borrow money. Buying something you can’t afford means spending more than you make. You’re not just borrowing from your lender; you are borrowing from your future self. Essentially, you are creating a time in the future in which you will need to spend less than you make so you can pay it back. The pattern of borrowing, spending more than you make, and then having to spend less than you make very quickly resembles a cycle. This is as true for a national economy as it is for an individual. Borrowing money sets a mechanical, predictable series of events into motion.
If you understand the game of Monopoly®, you can pretty well understand how credit cycles work on the level of a whole economy. Early in the game, people have a lot of cash and only a few properties, so it pays to convert your cash into property. As the game progresses and players acquire more and more houses and hotels, more and more cash is needed to pay the rents that are charged when you land on a property that has a lot of them. Some players are forced to sell their property at discounted prices to raise that cash. So early in the game, “property is
king” and later in the game, “cash is king.” Those who play the game best understand how to hold the right mix of property and cash as the game progresses.
Now, let’s imagine how this Monopoly® game would work if we allowed the bank to make loans and take deposits. Players would be able to borrow money to buy property, and, rather than holding their cash idly, they would deposit it at the bank to earn interest, which in turn would provide the bank with more money to lend. Let’s also imagine that players in this game could buy and sell properties from each other on credit (i.e., by promising to pay back the money with interest at a later date). If Monopoly® were played this way, it would provide an almost perfect model for the way our economy operates. The amount of debt-financed spending on hotels would quickly grow to multiples of the amount of money in existence. Down the road, the debtors who hold those hotels will become short on the cash they need to pay their rents and service their debt. The bank will also get into trouble as their depositors’ rising need for cash will cause them to withdraw it, even as more and more debtors are falling behind on their payments. If nothing is done to intervene, both banks and debtors will go broke and the economy will contract. Over time, as these cycles of expansion and contraction occur repeatedly, the conditions are created for a big, long-term debt crisis.
Lending naturally creates self-reinforcing upward movements that eventually reverse to create self-reinforcing downward movements that must reverse in turn. During the upswings, lending supports spending and investment, which in turn supports incomes and asset prices; increased incomes and asset prices support further borrowing and spending on goods and financial assets. The borrowing essentially lifts spending and incomes above the consistent productivity growth of the economy. Near the peak of the upward cycle, lending is based on the expectation that the above-trend growth will continue
indefinitely. But, of course, that can’t happen; eventually income will fall below the cost of the loans.
Economies whose growth is significantly supported by debt-financed building of fixed investments, real estate, and infrastructure are particularly susceptible to large cyclical swings because the fast rates of building those long-lived assets are not sustainable. If you need better housing and you build it, the incremental need to build more housing naturally declines. As spending on housing slows down, so does housing’s impact on growth. Let’s say you have been spending $10 million a year to build an office building (hiring workers, buying steel and concrete, etc.). When the building is finished, the spending will fall to $0 per year, as will the demand for workers and construction materials. From that point forward, growth, income, and the ability to service debt will depend on other demand. This type of cycle—where a strong growth upswing driven by debt- financed real estate, fixed investment, and infrastructure spending is followed by a downswing driven by a debt-challenged slowdown in demand—is very typical of emerging economies because they have so much building to do.
Contributing further to the cyclicality of emerging countries’ economies are changes in their competitiveness due to relative changes in their incomes. Typically, they have very cheap labor and bad infrastructure, so they build infrastructure, have an export boom, and experience rising incomes. But the rate of growth due to exports naturally slows as their income levels rise and their wage competitiveness relative to other countries declines. There are many examples of these kinds of cycles (i.e., Japan’s experience over the last 70 years).
In “bubbles,” the unrealistic expectations and reckless lending results in a critical mass of bad loans. At one stage or another, this becomes apparent to bankers and central bankers and the bubble begins to deflate. One classic warning sign that a bubble is coming is when an increasing amount of money is being borrowed to
make debt service payments, which of course compounds the borrowers’ indebtedness.
When money and credit growth are curtailed and/or higher lending standards are imposed, the rates of credit growth and spending slow and more debt service problems emerge. At this point, the top of the upward phase of the debt cycle is at hand. Realizing that credit growth is dangerously fast, the central banks tighten monetary policy to contain it, which often accelerates the decline (though it would have happened anyway, just a bit later). In either case, when the costs of debt service become greater than the amount that can be borrowed to finance spending, the upward cycle reverses. Not only does new lending slow down, but the pressure on debtors to make their payments is increased. The clearer it becomes that debtors are struggling, the less new lending there is. The slowdown in spending and investment that results slows down income growth even further, and asset prices decline.
When borrowers cannot meet their debt service obligations to lending institutions, those lending institutions cannot meet their obligations to their own creditors. Policy makers must handle this by dealing with the lending institutions first. The most extreme pressures are typically experienced by the lenders that are the most highly leveraged and that have the most concentrated exposures to failed borrowers. These lenders pose the biggest risks of creating knock-on effects for credit worthy buyers and across the economy. Typically, they are banks, but as credit systems have grown more dynamic, a broader set of lenders has emerged, such as insurance companies, non-bank trusts, broker-dealers, and even special purpose vehicles.
The two main long-term problems that emerge from these kinds of debt cycles are:
1. The losses arising from the expected debt service payments not being
made. When promised debt service payments can’t be made, that can lead to either smaller periodic payments and/or the writing down of the value of the debt (i.e., agreeing to accept less than was owed.) If you were expecting an annual debt service payment of 4 percent and it comes in at 2 percent or 0 percent, there is that shortfall for each year, whereas if the debt is marked down, that year’s loss would be much bigger (e.g., 50 percent).
2. The reduction of lending and the spending it was financing going forward. Even after a debt crisis is resolved, it is unlikely that the entities that borrowed too much can generate the same level of spending in the future that they had before the crisis. That has implications that must be considered.
Can Most Debt Crises Be Managed so There Aren’t Big Problems? Sometimes these cycles are moderate, like bumps in the road, and sometimes they are extreme, ending in crashes. In this study we examine ones that are extreme—i.e., all those in the last 100 years that produced declines in real GDP of more than 3 percent. Based on my examinations of them and the ways the levers available to policy makers work, I believe that it is possible for policy makers to manage them well in almost every case that the debts are denominated in a country’s own currency. That is because the flexibility that policy makers have allows them to spread out the harmful consequences in such ways that big debt problems aren’t really big problems. Most of the really terrible economic problems that debt crises have caused occurred before policy makers took steps to spread them out. Even the biggest debt crises in history (e.g., the 1930s Great Depression) were gotten past once the right adjustments were made. From my examination of these cases, the biggest risks are not from the debts themselves but from a) the failure of policy makers to do the right things, due to a lack of knowledge and/or lack of authority, and b) the political consequences of making adjustments that hurt some people in the process of helping others. It is from a desire to help reduce these risks that I have written this study.
Having said that, I want to reiterate that 1) when debts are denominated in foreign currencies rather than one’s own currency, it is much harder for a country’s policy makers to do the sorts of things that spread out the debt problems, and 2) the fact that debt crises can be well-managed does not mean that they are not extremely costly to some people.
The key to handling debt crises well lies in policy makers’ knowing how to use their levers well and having the authority that they need to do so, knowing at what rate per year the burdens will have to be spread out, and who will benefit and who will suffer and in what degree, so that the political and other consequences are acceptable.
There are four types of levers that policy makers can pull to bring debt and debt service levels down relative to the income and cash flow levels that are required to service them:
1. Austerity (i.e., spending less) 2. Debt defaults/restructurings 3. The central bank “printing money” and making purchases (or providing
guarantees) 4. Transfers of money and credit from those who have more than they need to
those who have less
Each one of their levers has different impacts on the economy. Some are inflationary and stimulate growth (e.g., “printing money”), while others are deflationary and help reduce debt burdens (e.g., austerity and defaults). The key to creating a “beautiful deleveraging” (a reduction in debt/income ratios accompanied by acceptable inflation and growth rates, which I explain later) lies in striking the right balance between them. In this happy scenario, debt-to- income ratios decline at the same time that economic activity and financial asset
prices improve, gradually bringing the nominal growth rate of incomes back above the nominal interest rate.
These levers shift around who benefits and who suffers, and over what amount of time. Policy makers are put in the politically difficult position of having to make those choices. As a result, they are rarely appreciated, even when they handle the debt crisis well.
The Template for the Archetypal Long- Term/Big Debt Cycle
The template that follows is based on my examination of 48 big debt cycles, which include all of the cases that led to real GDP falling by more than 3 percent in large countries (which is what I will call a depression). For clarity, I divided the affected countries into two groups: 1) Those that didn’t have much of their debt denominated in foreign currency and that didn’t experience inflationary depressions, and 2) those that had a significant amount of their debt denominated in foreign currency and did experience inflationary depressions. Since there was about a 75 percent correlation between the amounts of their foreign debts and the amounts of inflation that they experienced (which is not surprising, since having a lot of their debts denominated in foreign currency was a cause of their depressions being inflationary), it made sense to group those that had more foreign currency debt with those that had inflationary depressions.
Typically debt crises occur because debt and debt service costs rise faster than the incomes that are needed to service them, causing a deleveraging. While the central bank can alleviate typical debt crises by lowering real and nominal interest rates, severe debt crises (i.e., depressions) occur when this is no longer possible. Classically, a lot of short-term debt cycles (i.e., business cycles) add up to a long-term debt cycle, because each short-term cyclical high and each short- term cyclical low is higher in its debt-to-income ratio than the one before it, until the interest rate reductions that helped fuel the expansion in debt can no longer continue. The chart below shows the debt and debt service burden (both principal and interest) in the US since 1910. You will note how the interest
payments remain flat or go down even when the debt goes up, so that the rise in debt service costs is not as great as the rise in debt. That is because the central bank (in this case, the Federal Reserve) lowers interest rates to keep the debt- financed expansion going until they can’t do it any more (because the interest rate hits 0 percent). When that happens, the deleveraging begins.
While the chart gives a good general picture, I should make clear that it is inadequate in two respects: 1) it doesn’t convey the differences between the various entities that make up these total numbers, which are very important to understand, and 2) it just shows what is called debt, so it doesn’t reflect liabilities such as pension and health care obligations, which are much larger. Having this more granular perspective is very important in gauging a country’s vulnerabilities, though for the most part such issues are beyond the scope of this book.
Our Examination of the Cycle
In developing the template, we will focus on the period leading up to the depression, the depression period itself, and the deleveraging period that follows the bottom of the depression. As there are two broad types of big debt crises— deflationary ones and inflationary ones (largely depending on whether a country has a lot of foreign currency debt or not)—we will examine them separately.
The statistics reflected in the charts of the phases were derived by averaging 21 deflationary debt cycle cases and 27 inflationary debt cycle cases, starting five years before the bottom of the depression and continuing for seven years after it.
Notably long-term debt cycles appear similar in many ways to short-term debt cycles, except that they are more extreme, both because the debt burdens are higher and the monetary policies that can address them are less effective. For the most part, short-term debt cycles produce bumps—mini-booms and recessions— while big long-term ones produce big booms and busts. Over the last century, the US has gone through a long-term debt crisis twice—once during the boom of the 1920s and the Great Depression of the 1930s, and again during the boom of the early 2000s and the financial crisis starting in 2008.
In the short-term debt cycle, spending is constrained only by the willingness of lenders and borrowers to provide and receive credit. When credit is easily available, there’s an economic expansion. When credit isn’t easily available, there’s a recession. The availability of credit is controlled primarily by the central bank. The central bank is generally able to bring the economy out of a recession by easing rates to stimulate the cycle anew. But over time, each bottom and top of the cycle finishes with more economic activity than the previous
cycle, and with more debt. Why? Because people push it—they have an inclination to borrow and spend more instead of paying back debt. It’s human nature. As a result, over long periods of time, debts rise faster than incomes. This creates the long-term debt cycle.
During the upswing of the long-term debt cycle, lenders extend credit freely even as people become more indebted. That’s because the process is self- reinforcing on the upside—rising spending generates rising incomes and rising net worths, which raises borrowers’ capacities to borrow, which allows more buying and spending, etc. Most everyone is willing to take on more risk. Quite often new types of financial intermediaries and new types of financial instruments develop that are outside the supervision and protection of regulatory authorities. That puts them in a competitively attractive position to offer higher returns, take on more leverage, and make loans that have greater liquidity or credit risk. With credit plentiful, borrowers typically spend more than is sustainable, giving them the appearance of being prosperous. In turn, lenders, who are enjoying the good times, are more complacent than they should be. But debts can’t continue to rise faster than the money and income that is necessary to service them forever, so they are headed toward a debt problem.
When the limits of debt growth relative to income growth are reached, the process works in reverse. Asset prices fall, debtors have problems servicing their debts, and investors get scared and cautious, which leads them to sell, or not roll over, their loans. This, in turn, leads to liquidity problems, which means that people cut back on their spending. And since one person’s spending is another person’s income, incomes begin to go down, which makes people even less creditworthy. Asset prices fall, further squeezing banks, while debt repayments continue to rise, making spending drop even further. The stock market crashes and social tensions rise along with unemployment, as credit and cash-starved companies reduce their expenses. The whole thing starts to feed on itself the
other way, becoming a vicious, self-reinforcing contraction that’s not easily corrected. Debt burdens have simply become too big and need to be reduced. Unlike in recessions, when monetary policies can be eased by lowering interest rates and increasing liquidity, which in turn increase the capacities and incentives to lend, interest rates can’t be lowered in depressions. They are already at or near zero and liquidity/money can’t be increased by ordinary measures.
This is the dynamic that creates long-term debt cycles. It has existed for as long as there has been credit, going back to before Roman times. Even the Old Testament described the need to wipe out debt once every 50 years, which was called the Year of Jubilee. Like most dramas, this one both arises and transpires in ways that have reoccurred throughout history.
Remember that money serves two purposes: it is a medium of exchange and a store hold of wealth. And because it has two purposes, it serves two masters: 1) those who want to obtain it for “life’s necessities,” usually by working for it, and 2) those who have stored wealth tied to its value. Throughout history these two groups have been called different things—e.g., the first group has been called workers, the proletariat, and “the have-nots,” and the second group has been called capitalists, investors, and “the haves.” For simplicity, we will call the first group proletariat-workers and the second group capitalists-investors. Proletariat- workers earn their money by selling their time and capitalists-investors earn their money by “lending” others the use of their money in exchange for either a) a promise to repay an amount of money that is greater than the loan (which is a debt instrument), or b) a piece of ownership in the business (which we call “equity” or “stocks”) or a piece of another asset (e.g., real estate). These two groups, along with the government (which sets the rules), are the major players in this drama. While generally both groups benefit from borrowing and lending,
sometimes one gains and one suffers as a result of the transaction. This is especially true for debtors and creditors.
One person’s financial assets are another’s financial liabilities (i.e., promises to deliver money). When the claims on financial assets are too high relative to the money available to meet them, a big deleveraging must occur. Then the free- market credit system that finances spending ceases to work well, and typically works in reverse via a deleveraging, necessitating the government to intervene in a big way as the central bank becomes a big buyer of debt (i.e., lender of last resort) and the central government becomes a redistributor of spending and wealth. At such times, there needs to be a debt restructuring in which claims on future spending (i.e., debt) are reduced relative to what they are claims on (i.e., money).
This fundamental imbalance between the size of the claims on money (debt) and the supply of money (i.e., the cash flow that is needed to service the debt) has occurred many times in history and has always been resolved via some combination of the four levers I previously described. The process is painful for all of the players, sometimes so much so that it causes a battle between the proletariat-workers and the capitalist-investors. It can get so bad that lending is impaired or even outlawed. Historians say that the problems that arose from credit creation were why usury (lending money for interest) was considered a sin in both Catholicism and Islam.
In this study we will examine big debt cycles that produce big debt crises, exploring how they work and how to deal with them well. But before we begin, I want to clarify the differences between the two main types: deflationary and inflationary depressions.
In deflationary depressions, policy makers respond to the initial economic contraction by lowering interest rates. But when interest rates reach about 0
2
percent, that lever is no longer an effective way to stimulate the economy. Debt restructuring and austerity dominate, without being balanced by adequate stimulation (especially money printing and currency depreciation). In this phase, debt burdens (debt and debt service as a percent of income) rise, because incomes fall faster than restructuring, debt paydowns reduce the debt stock, and many borrowers are required to rack up still more debts to cover those higher interest costs. As noted, deflationary depressions typically occur in countries where most of the unsustainable debt was financed domestically in local currency, so that the eventual debt bust produces forced selling and defaults, but not a currency or a balance of payments problem. Inflationary depressions classically occur in countries that are reliant on foreign capital flows and so have built up a significant amount of debt denominated in foreign currency that can’t be monetized (i.e., bought by money printed by the central bank). When those foreign capital flows slow, credit creation turns into credit contraction. In an inflationary deleveraging, capital withdrawal dries up lending and liquidity at the same time that currency declines produce inflation. Inflationary depressions in which a lot of debt is denominated in foreign currency are especially difficult to manage because policy makers’ abilities to spread out the pain are more limited.
We will begin with deflationary depressions.
1. I’m so debt adverse that I’ve hardly had any debt in any form, even when I bought my first house. When I built Bridgewater, it was without debt, and I’m still a keen saver.
2. Throughout the Middle Ages, Christians could generally not legally charge interest to other Christians. This is one reason why Jews played a large part in the development of trade, as they lent money for business ventures and financed voyages. But Jews were also the holders of the loans that debtors sometimes could not repay. Many historical instances of violence against Jews were driven by debt crises.
The Phases of the Classic Deflationary Debt Cycle
The chart below illustrates the seven stages of an archetypal long-term debt cycle, by tracking the total debt of the economy as a percentage of the total income of the economy (GDP) and the total amount of debt service payments relative to GDP over a period of 12 years.
Throughout this section, I’ll include similar “archetype” charts that are built by averaging the deflationary deleveraging cases.
1) The Early Part of the Cycle In the early part of the cycle, debt is not growing faster than incomes, even though debt growth is strong. That is because debt growth is being used to finance activities that produce fast income growth. For instance, borrowed money may go toward expanding a business and making it more productive,
3
supporting growth in revenues. Debt burdens are low and balance sheets are healthy, so there is plenty of room for the private sector, government, and banks to lever up. Debt growth, economic growth, and inflation are neither too hot nor too cold. This is what is called the “Goldilocks” period.
2) The Bubble In the first stage of the bubble, debts rise faster than incomes, and they produce accelerating strong asset returns and growth. This process is generally self- reinforcing because rising incomes, net-worths, and asset values raise borrowers’ capacities to borrow. This happens because lenders determine how much they can lend on the basis of the borrowers’ 1) projected income/cash flows to service the debt, 2) net worth/collateral (which rises as asset prices rise), and 3) their own capacities to lend. All of these rise together. Though this set of conditions is not sustainable because the debt growth rates are increasing faster than the incomes that will be required to service them, borrowers feel rich, so they spend more than they earn and buy assets at high prices with leverage. Here’s one example of how that happens:
Suppose you earn $50,000 a year and have a net worth of $50,000. You have the capacity to borrow $10,000 per year, so you could spend $60,000 per year for a number of years, even though you only earn $50,000. For an economy as a whole, increased borrowing and spending can lead to higher incomes, and rising stock valuations and other asset values, giving people more collateral to borrow against. People then borrow more and more, but as long as the borrowing drives growth, it is affordable.
In this up-wave part of the long-term debt cycle, promises to deliver money (i.e., debt burdens) rise relative to both the supply of money in the overall economy and the amount of money and credit debtors have coming in (via incomes,
borrowing, and sales of assets). This up-wave typically goes on for decades, with variations primarily due to central banks’ periodic tightenings and easings of credit. These are short-term debt cycles, and a bunch of them generally add up to a long-term debt cycle.
A key reason the long-term debt cycle can be sustained for so long is that central banks progressively lower interest rates, which raises asset prices and, in turn, people’s wealth, because of the present value effect that lowering interest rates has on asset prices. This keeps debt service burdens from rising, and it lowers the monthly payment cost of items bought on credit. But this can’t go on forever. Eventually the debt service payments become equal to or larger than the amount debtors can borrow, and the debts (i.e., the promises to deliver money) become too large in relation to the amount of money in existence there is to give. When promises to deliver money (i.e., debt) can’t rise any more relative to the money and credit coming in, the process works in reverse and deleveraging begins. Since borrowing is simply a way of pulling spending forward, the person spending $60,000 per year and earning $50,000 per year has to cut his spending to $40,000 for as many years as he spent $60,000, all else being equal.
Though a bit of an oversimplification, this is the essential dynamic that drives the inflating and deflating of a bubble.
The Start of a Bubble: The Bull Market Bubbles usually start as over-extrapolations of justified bull markets. The bull markets are initially justified because lower interest rates make investment assets, such as stocks and real estate, more attractive so they go up, and economic conditions improve, which leads to economic growth and corporate profits, improved balance sheets, and the ability to take on more debt—all of which make the companies worth more.
As assets go up in value, net worths and spending/income levels rise. Investors, business people, financial intermediaries, and policy makers increase their confidence in ongoing prosperity, which supports the leveraging-up process. The boom also encourages new buyers who don’t want to miss out on the action to enter the market, fueling the emergence of a bubble. Quite often, uneconomic lending and the bubble occur because of implicit or explicit government guarantees that encourage lending institutions to lend recklessly.
As new speculators and lenders enter the market and confidence increases, credit standards fall. Banks lever up and new types of lending institutions that are largely unregulated develop (these nonbank lending institutions are referred to collectively as a “shadow banking” system). These shadow banking institutions are typically less under the blanket of government protections. At these times, new types of lending vehicles are frequently invented and a lot of financial engineering takes place.
The lenders and the speculators make a lot of fast, easy money, which reinforces the bubble by increasing the speculators’ equity, giving them the collateral they need to secure new loans. At the time, most people don’t think that is a problem; to the contrary, they think that what is happening is a reflection and confirmation of the boom. This phase of the cycle typically feeds on itself. Taking stocks as an example, rising stock prices lead to more spending and investment, which raises earnings, which raises stock prices, which lowers credit spreads and encourages increased lending (based on the increased value of collateral and higher earnings), which affects spending and investment rates, etc. During such times, most people think the assets are a fabulous treasure to own—and consider anyone who doesn’t own them to be missing out. As a result of this dynamic, all sorts of entities build up long positions. Large asset-liability mismatches increase in the forms of a) borrowing short-term to lend long-term, b) taking on liquid liabilities to invest in illiquid assets, and c) investing in riskier debt or
other risky assets with money borrowed from others, and/or d) borrowing in one currency and lending in another, all to pick up a perceived spread. All the while, debts rise fast and debt service costs rise even faster. The charts below paint the picture.
In markets, when there’s a consensus, it gets priced in. This consensus is also typically believed to be a good rough picture of what’s to come, even though history has shown that the future is likely to turn out differently than expected. In other words, humans by nature (like most species) tend to move in crowds and weigh recent experience more heavily than is appropriate. In these ways, and because the consensus view is reflected in the price, extrapolation tends to occur.
At such times, increases in debt-to-income ratios are very rapid. The above chart shows the archetypal path of debt as a percent of GDP for the deflationary deleveragings we averaged. The typical bubble sees leveraging up at an average rate of 20 to 25 percent of GDP over three years or so. The blue line depicts the arc of the long-term debt cycle in the form of the total debt of the economy divided by the total income of the economy as it passes through its various phases; the red line charts the total amount of debt service payments relative to the total amount of income.
Bubbles are most likely to occur at the tops in the business cycle, balance of payments cycle, and/or long-term debt cycle. As a bubble nears its top, the economy is most vulnerable, but people are feeling the wealthiest and the most bullish. In the cases we studied, total debt-to-income levels averaged around 300 percent of GDP. To convey a few rough average numbers, below we show some key indications of what the archetypal bubble looks like:
The Role of Monetary Policy In many cases, monetary policy helps inflate the bubble rather than constrain it. This is especially true when inflation and growth are both good and investment returns are great. Such periods are typically interpreted to be a productivity boom that reinforces investor optimism as they leverage up to buy investment assets. In such cases, central banks, focusing on inflation and growth, are often reluctant to adequately tighten money. This is what happened in Japan in the late 1980s, and in much of the world in the late 1920s and mid-2000s.
This is one of the biggest problems with most central bank policies—i.e., because central bankers target either inflation or inflation and growth and don’t target the management of bubbles, the debt growth that they enable can go to finance the creation of bubbles if inflation and real growth don’t appear to be too strong. In my opinion it’s very important for central banks to target debt growth with an eye toward keeping it at a sustainable level—i.e., at a level where the growth in income is likely to be large enough to service the debts regardless of what credit is used to buy. Central bankers sometimes say that it is too hard to spot bubbles and that it’s not their role to assess and control them—that it is their job to control inflation and growth. But what they control is money and credit, and when that money and credit goes into debts that can’t be paid back, that has
4
huge implications for growth and inflation down the road. The greatest depressions occur when bubbles burst, and if the central banks that are producing the debts that are inflating them won’t control them, then who will? The economic pain of allowing a large bubble to inflate and then burst is so high that it is imprudent for policy makers to ignore them, and I hope their perspective will change.
While central banks typically do tighten money somewhat and short rates rise on average when inflation and growth start to get too hot, typical monetary policies are not adequate to manage bubbles, because bubbles are occurring in some parts of the economy and not others. Thinking about the whole economy, central banks typically fall behind the curve during such periods, and borrowers are not yet especially squeezed by higher debt-service costs. Quite often at this stage, their interest payments are increasingly being covered by borrowing more rather than by income growth—a clear sign that the trend is unsustainable.
All this reverses when the bubble pops and the same linkages that inflated the bubble make the downturn self-reinforcing. Falling asset prices decrease both the equity and collateral values of leveraged speculators, which causes lenders to pull back. This forces speculators to sell, driving down prices even more. Also, lenders and investors “run” (i.e., withdraw their money) from risky financial intermediaries and risky investments, causing them to have liquidity problems. Typically, the affected market or markets are big enough and leveraged enough that the losses on the accumulated debt are systemically threatening, which is to say that they threaten to topple the entire economy.
Spotting Bubbles While the particulars may differ across cases (e.g., the size of the bubble; whether it’s in stocks, housing, or some other asset ; how exactly the bubble pops; and so on), the many cases of bubbles are much more similar than they are different, and each is a result of logical cause-and-effect relationships that can be studied and understood. If one holds a strong mental map of how bubbles form, it becomes much easier to identify them.
To identify a big debt crisis before it occurs, I look at all the big markets and see which, if any, are in bubbles. Then I look at what’s connected to them that would be affected when they pop. While I won’t go into exactly how it works here, the most defining characteristics of bubbles that can be measured are:
1. Prices are high relative to traditional measures 2. Prices are discounting future rapid price appreciation from these high levels 3. There is broad bullish sentiment 4. Purchases are being financed by high leverage 5. Buyers have made exceptionally extended forward purchases (e.g., built
inventory, contracted for supplies, etc.) to speculate or to protect
5
themselves against future price gains 6. New buyers (i.e., those who weren’t previously in the market) have entered
the market 7. Stimulative monetary policy threatens to inflate the bubble even more (and
tight policy to cause its popping)
As you can see in the table below, which is based on our systematic measures, most or all of these indications were present in past bubbles. (N/A indicates inadequate data.)
At this point I want to emphasize that it is a mistake to think that any one metric can serve as an indicator of an impending debt crisis. The ratio of debt to income for the economy as a whole, or even debt service payments to income for the economy as a whole, which is better, are useful but ultimately inadequate measures. To anticipate a debt crisis well, one has to look at the specific debt- service abilities of the individual entities, which are lost in these averages. More specifically, a high level of debt or debt service to income is less problematic if the average is well distributed across the economy than if it is concentrated— especially if it is concentrated in key entities.
3) The Top
When prices have been driven by a lot of leveraged buying and the market gets fully long, leveraged, and overpriced, it becomes ripe for a reversal. This reflects a general principle: When things are so good that they can’t get better—yet everyone believes that they will get better—tops of markets are being made.
While tops are triggered by different events, most often they occur when the central bank starts to tighten and interest rates rise. In some cases the tightening is brought about by the bubble itself, because growth and inflation are rising while capacity constraints are beginning to pinch. In other cases, the tightening is externally driven. For example, for a country that has become reliant on borrowing from external creditors, the pulling back of lending due to exogenous causes will lead to liquidity tightening. A tightening of monetary policy in the currency in which debts are denominated can be enough to cause foreign capital to pull back. This can happen for reasons unrelated to conditions in the domestic economy (e.g., cyclical conditions in a reserve currency country leads to a tightening in liquidity in that currency, or a financial crisis results in a pullback of capital, etc.). Also, a rise in the currency the debt is in relative to the currency incomes are in can cause an especially severe squeeze. Sometimes unanticipated shortfalls in cash flows due to any number of reasons can trigger the debt crises.
Whatever the cause of the debt-service squeeze, it hurts asset prices (e.g., stock prices), which has a negative “wealth effect” as lenders begin to worry that they might not be able to get their cash back from those they lent it to. Borrowers are squeezed as an increasing share of their new borrowing goes to pay debt service and/or isn’t rolled over and their spending slows down. This is classically the result of people buying investment assets at high prices with leverage, based on overly optimistic assumptions about future cash flow. Typically, these types of credit/debt problems start to emerge about half a year ahead of the peak in the economy, at first in its most vulnerable and frothy pockets. The riskiest debtors start to miss payments, lenders begin to worry, credit spreads start to tick up, and
6
risky lending slows. Runs from risky assets to less risky assets pick up, contributing to a broadening of the contraction.
Typically, in the early stages of the top, the rise in short rates narrows or eliminates the spread with long rates (i.e., the extra interest rate earned for lending long term rather than short term), lessening the incentive to lend relative to the incentive to hold cash. As a result of the yield curve being flat or inverted (i.e., long-term interest rates are at their lowest relative to short-term interest rates), people are incentivized to move to cash just before the bubble pops, slowing credit growth and causing the previously described dynamic.
Early on in the top, some parts of the credit system suffer, but others remain robust, so it isn’t clear that the economy is weakening. So while the central bank is still raising interest rates and tightening credit, the seeds of the recession are being sown. The fastest rate of tightening typically comes about five months prior to the top of the stock market. The economy is then operating at a high rate, with demand pressing up against the capacity to produce. Unemployment is normally at cyclical lows and inflation rates are rising. The increase in short- term interest rates makes holding cash more attractive, and it raises the interest rate used to discount the future cash flows of assets, weakening riskier asset prices and slowing lending. It also makes items bought on credit de facto more
expensive, slowing demand. Short rates typically peak just a few months before the top in the stock market.
The more leverage that exists and the higher the prices, the less tightening it takes to prick the bubble and the bigger the bust that follows. To understand the magnitude of the downturn that is likely to occur, it is less important to understand the magnitude of the tightening than it is to understand each particular sector’s sensitivity to tightening and how losses will cascade. These pictures are best seen by looking at each of the important sectors of the economy and each of the big players in these sectors rather than at economy-wide averages.
In the immediate postbubble period, the wealth effect of asset price movements has a bigger impact on economic growth rates than monetary policy does. People tend to underestimate the size of this effect. In the early stages of a bubble bursting, when stock prices fall and earnings have not yet declined, people mistakenly judge the decline to be a buying opportunity and find stocks cheap in relation to both past earnings and expected earnings, failing to account for the amount of decline in earnings that is likely to result from what’s to come. But the reversal is self-reinforcing. As wealth falls first and incomes fall later, creditworthiness worsens, which constricts lending activity, which hurts
spending and lowers investment rates while also making it less appealing to borrow to buy financial assets. This in turn worsens the fundamentals of the asset (e.g., the weaker economic activity leads corporate earnings to chronically disappoint), leading people to sell and driving down prices further. This has an accelerating downward impact on asset prices, income, and wealth.
4) The “Depression”
In normal recessions (when monetary policy is still effective), the imbalance between the amount of money and the need for it to service debt can be rectified by cutting interest rates enough to 1) produce a positive wealth effect, 2) stimulate economic activity, and 3) ease debt-service burdens. This can’t happen in depressions, because interest rates can’t be cut materially because they have either already reached close to 0 percent or, in cases where currency outflows and currency weaknesses are great, the floor on interest rates is higher because of credit or currency risk considerations.
This is precisely the formula for a depression. As shown, this happened in the early stage of both the 1930–32 depression and the 2008–09 depression. In well managed cases, like the US in 2007–08, the Fed lowered rates very quickly and
then, when that didn’t work, moved on to alternative means of stimulating, having learned from its mistakes in the 1930s when the Fed was slower to ease and even tightened at times to defend the dollar’s peg to gold.
The chart below shows the sharp lowering of interest rates toward 0 percent for the average of the 21 deflationary debt crises that we looked at.
As the depression begins, debt defaults and restructurings hit the various players, especially leveraged lenders (e.g., banks), like an avalanche. Both lenders’ and depositors’ justified fears feed on themselves, leading to runs on financial institutions that typically don’t have the cash to meet them unless they are under the umbrella of government protections. Cutting interest rates doesn’t work adequately because the floors on risk-free rates have already been hit and
because as credit spreads rise, the interest rates on risky loans go up, making it difficult for those debts to be serviced. Interest rate cuts also don’t do much to help lending institutions that have liquidity problems and are suffering from runs. At this phase of the cycle, debt defaults and austerity (i.e., the forces of deflation) dominate, and are not sufficiently balanced with the stimulative and inflationary forces of printing money to cover debts (i.e., debt monetization).
With investors unwilling to continue lending and borrowers scrambling to find cash to cover their debt payments, liquidity—i.e., the ability to sell investments for money—becomes a major concern. As an illustration, when you own a $100,000 debt instrument, you presume that you will be able to exchange it for $100,000 in cash and, in turn, exchange the cash for $100,000 worth of goods and services. However since the ratio of financial assets to money is high, when a large number of people rush to convert their financial assets into money and buy goods and services in bad times, the central bank either has to provide the liquidity that’s needed by printing more money or allow a lot of defaults.
The depression can come from, or cause, either solvency problems or cash-flow problems. Usually a lot of both types of problems exist during this phase. A solvency problem means that, according to accounting and regulatory rules, the entity does not have enough equity capital to operate—i.e., it is “broke” and must be shut down. So, the accounting laws have a big impact on the severity of the debt problem at this moment. A cash-flow problem means that an entity doesn’t have enough cash to meet its needs, typically because its own lenders are taking money away from it—i.e., there is a “run.” A cash-flow problem can occur even when the entity has adequate capital because the equity is in illiquid assets. Lack of cash flow is an immediate and severe problem—and as a result, the trigger and main issue of most debt crises.
Each kind of problem requires a different approach. If a solvency problem exists (i.e., the debtor doesn’t have enough equity capital), it has an
accounting/regulatory problem that can be dealt with by either a) providing enough equity capital or b) changing the accounting/regulatory rules, which hides the problem. Governments can do this directly through fiscal policy or indirectly through clever monetary policies if the debt is in their own currency. Similarly, if a cash-flow problem exists, fiscal and/or monetary policy can provide either cash or guarantees that resolve it.
A good example of how these forces are relevant is highlighted by the differences between the debt/banking crises of the 1980s and 2008. In the 1980s, there was not as much mark-to-market accounting (because the crisis involved loans that weren’t traded every day in public markets), so the banks were not as “insolvent” as they were in 2008. With more mark-to-market accounting in 2008, the banks required capital injections and/or guarantees to improve their balance sheets. Both crises were successfully managed, though the ways they were managed had to be different.
Going into the “depression” phase of the cycle (by which I mean the severe contraction phase) some protections learned from past depressions (e.g., bank- deposit insurance, the ability to provide lender-of-last-resort financial supports and guarantees and to inject capital into systemically important institutions or nationalize them) are typically in place and are helpful, but they are rarely adequate, because the exact nature of the debt crisis hasn’t been well thought through. Typically, quite a lot of lending has taken place in the relatively unregulated “shadow banking system,” or in new instruments that have unanticipated risks and inadequate regulations. What happens in response to these new realities depends on the capabilities of the policy makers in the decision roles and the freedom of the system to allow them to do what is best.
Some people mistakenly think that depressions are psychological: that investors move their money from riskier investments to safer ones (e.g., from stocks and high-yield lending to government bonds and cash) because they’re scared, and
that the economy will be restored if they can only be coaxed into moving their money back into riskier investments. This is wrong for two reasons: First, contrary to popular belief, the deleveraging dynamic is not primarily psychological. It is mostly driven by the supply and demand of, and the relationships between, credit, money, and goods and services—though psychology of course also does have an effect, especially in regard to the various players’ liquidity positions. Still, if everyone went to sleep and woke up with no memory of what had happened, we would be in the same position, because debtors’ obligations to deliver money would be too large relative to the money they are taking in. The government would still be faced with the same choices that would have the same consequences, and so on.
Related to this, if the central bank produces more money to alleviate the shortage, it will cheapen the value of money, making a reality of creditors’ worries about being paid back an amount of money that is worth less than what they loaned. While some people think that the amount of money in existence remains the same and simply moves from riskier assets to less risky ones, that’s not true. Most of what people think is money is really credit, and credit does appear out of thin air during good times and then disappear at bad times. For example, when you buy something in a store on a credit card, you essentially do so by saying, “I promise to pay.” Together you and the store owner create a credit asset and a credit liability. So where do you take the money from? Nowhere. You created credit. It goes away in the same way. Suppose the store owner rightly believes that you and others won’t pay the credit card company and that the credit card company won’t pay him. Then he correctly believes that the credit “asset” he has isn’t really there. It didn’t go somewhere else; it’s simply gone.
As this implies, a big part of the deleveraging process is people discovering that much of what they thought of as their wealth was merely people’s promises to
give them money. Now that those promises aren’t being kept, that wealth no longer exists. When investors try to convert their investments into money in order to raise cash, they test their ability to get paid, and in cases where it fails, panic-induced “runs” and sell-offs of securities occur. Naturally those who experience runs, especially banks (though this is true of most entities that rely on short-term funding), have problems raising money and credit to meet their needs, so debt defaults cascade.
Debt defaults and restructurings hit people, especially leveraged lenders (e.g., banks), and fear cascades through the system. These fears feed on themselves and lead to a scramble for cash that results in a shortage (i.e., a liquidity crisis). The dynamic works like this: Initially, the money coming in to debtors via incomes and borrowing is not enough to meet the debtors’ obligations; assets need to be sold and spending needs to be cut in order to raise cash. This leads asset values to fall, which reduces the value of collateral, and in turn reduces incomes. Since borrowers’ creditworthiness is judged by both a) the values of their assets/collaterals in relation to their debts (i.e., their net worth) and b) the sizes of their incomes relative to the sizes of their debt-service payments, and since both their net worth and their income fall faster than their debts, borrowers become less creditworthy and lenders more reluctant to lend. This goes on in a self-reinforcing manner.
The depression phase is dominated by the deflationary forces of debt reduction (i.e., defaults and restructurings) and austerity occurring without material efforts to reduce debt burdens by printing money. Because one person’s debts are another’s assets, the effect of aggressively cutting the value of those assets can be to greatly reduce the demand for goods, services, and investment assets. For a write-down to be effective, it must be large enough to allow the debtor to service the restructured loan. If the write-down is 30 percent, then the creditor’s assets are reduced by that much. If that sounds like a lot, it’s actually much more. Since
most lenders are leveraged (e.g., they borrow to buy assets), the impact of a 30 percent write-down on their net worth can be much greater. For example, the creditor who is leveraged 2:1 would experience a 60 percent decline in his net worth (i.e., their assets are twice their net worth, so the decline in asset value has twice the impact). Since banks are typically leveraged about 12:1 or 15:1, that picture is obviously devastating for them and for the economy as a whole.
Even as debts are written down, debt burdens rise as spending and incomes fall. Debt levels also rise relative to net worth, as shown in the chart below. As debt- to-income and debt-to-net-worth ratios go up and the availability of credit goes down, naturally the credit contraction becomes self-reinforcing on the downside.
The capitalist-investor class experiences a tremendous loss of “real” wealth during depressions because the value of their investment portfolios collapses (declines in equity prices are typically around 50 percent), their earned incomes fall, and they typically face higher tax rates. As a result, they become extremely defensive. Quite often, they are motivated to move their money out of the country (which contributes to currency weakness), dodge taxes, and seek safety in liquid, noncredit-dependent investments (e.g., low-risk government bonds, gold, or cash).
Of course, the real economy as well as the financial economy suffers. With monetary policy constrained, the uncontrolled credit contraction produces an
7
economic and social catastrophe. Workers suffer as incomes collapse and job losses are severe. Hard-working people who once were able to provide for their families lose the opportunity to have meaningful work and suddenly become either destitute or dependent. Homes are lost because owners can no longer afford to pay their mortgages, retirement accounts are wiped out, and savings for college are lost. These conditions can persist for many years if policy makers don’t offset the depression’s deflationary forces with sufficient monetary stimulation of a new form.
Managing Depressions As mentioned earlier, the policies that reduce debt burdens fall under four broad categories: 1) austerity, 2) debt defaults/restructurings, 3) debt monetization/money printing, and 4) wealth transfers (i.e., from the haves to the have-nots). By using these kinds of levers well, policy makers can mitigate the worst effects of a depression and manage both the failed lenders and borrowers and the economic conditions. But it’s important to recognize that each of these levers has different impacts on the economy and creditworthiness. The key is to get the mix right, so that deflationary and depressive forces are balanced with inflationary and stimulative ones.
Policy makers typically get the mix between austerity, money printing, and redistribution wrong initially. Taxpayers are understandably angry at the debtors and at the financial institutions whose excesses caused the debt crisis, and don’t want the government (i.e., their taxes) to bail them out. And policy makers justifiably believe that debt excesses will happen again if lenders and borrowers don’t suffer the downsides of their actions (which is called the “moral hazard” problem). For all these reasons, policy makers are usually reluctant and slow to provide government supports, and the debt contraction and the agony it produces increase quickly. But the longer they wait to apply stimulative remedies to the
8
mix, the uglier the deleveraging becomes. Eventually they choose to provide a lot of guarantees, print a lot of money, and monetize a lot of debt, which lifts the economy into a reflationary deleveraging. If they do these things and get the mix right quickly, the depression is much more likely to be relatively short-lived (like the short period of “depression” following the US crisis in 2008). If they don’t, the depression is usually prolonged (like the Great Depression in the 1930s, or Japan’s “lost decade” following its bubble in the late 1980s).
To reiterate, the two biggest impediments to managing a debt crisis are: a) the failure to know how to handle it well and b) politics or statutory limitations on the powers of policy makers to take the necessary actions. In other words, ignorance and a lack of authority are bigger problems than debts themselves. While being a successful investment manager is hard, it’s not nearly as hard as being a successful economic policy maker. We investors only have to understand how the economic machine works and anticipate what will happen next. Policy makers have to do that, plus make everything turn out well—i.e., they have to know what should be done while navigating through all the political impediments that make it so hard to get it done. To do that requires a lot of smarts, a willingness to fight, and political savvy—i.e., skills and heroism—and sometimes even with all those things, the constraints under which they work still prevent them from being successful.
Below I’ll walk through each of the four levers and how they are typically used in the depression phase.
Austerity In the depression phase, policy makers typically try austerity because that’s the obvious thing to do. It’s natural to want to let those who got themselves and others in trouble to bear the costs. The problem is that even deep austerity
8
doesn’t bring debt and income back into balance. When spending is cut, incomes are also cut, so it takes an awful lot of painful spending cuts to make significant reductions in the debt/income ratios.
As the economy contracts, government revenues typically fall. At the same time, demands on the government increase. As a result, deficits typically increase. Seeking to be fiscally responsible, at this point governments tend to raise taxes.
Both moves are big mistakes.
“Printing Money” to Stop the Bleeding and Stimulate the Economy Quite often “runs” on lending institutions occur, especially those that aren’t protected by government guarantees. That puts the central bank and the central government in the position of having to decide which depositors/lenders should be protected from losses and which should be allowed to sustain them, and which institutions are systemically important and should be saved—and how to do these things in a way that maximizes the safety of the financial/economic system while minimizing costs to the government/taxpayers. At such times all sorts of guarantees are offered to systemically critical financial institutions—and quite often some of these institutions are nationalized. Typically there are a lot of laws and politics that affect how quickly and how well this is done. Some of the money that is needed comes from the government (i.e., it is appropriated through the budget process) and some from the central banks (by “printing”). Governments inevitably do both, though in varying degrees. In addition to providing money to some essential banks, governments also typically provide money to some nonbank entities they consider essential.
Next, they must ease the credit crunch and stimulate the overall economy. Since the government is likely having trouble raising funds through taxation and borrowing, central banks are forced to choose between “printing” still more money to buy their governments’ debts or allowing their governments and their private sector to compete for the limited supply of money, which will only tighten money further. Inevitably, they choose to print.
Typically, though not necessarily, these moves come in progressively larger doses as more modest initial attempts fail to rectify the imbalance and reverse the deleveraging process. However, those early efforts do typically cause temporary periods of relief that are manifest in bear-market rallies in financial assets and increased economic activity. During the Great Depression there were six big rallies in the stock market (of between 16 percent and 48 percent) in a bear market that declined a total of 89 percent. All of those rallies were triggered by government actions that were intended to reduce the fundamental imbalance. When they are managed well, those shifts in policies to “print money,” buy assets, and provide guarantees are what moves the debt cycle from its depression/“ugly deleveraging” phase to its expansion/“beautiful deleveraging” phase. The chart below shows how this “money printing” happened in the US in the 1930s and again after 2008.
While highly stimulative monetary policy is a critical part of a deleveraging, it is typically not sufficient. When risks emerge that systemically important institutions will fail, policy makers must take steps to keep these entities running. They must act immediately to:
Curtail panic and guarantee liabilities. Governments can increase guarantees on deposits and debt issuance. Central banks can provide systemically important institutions (i.e., institutions whose failures would threaten the ongoing operating of the financial system and/or that of the economy) with injections of money. Occasionally, governments can force liquidity to remain in the banking system by imposing deposit freezes, which is generally undesirable because it intensifies the panic, but is sometimes necessary because there is no other way of providing that money/liquidity. Provide liquidity. When private credit is contracting and liquidity is tight, the central bank can ensure that sufficient liquidity is provided to the financial system by lending against a widening range of collateral or to an increasingly wide range of financial institutions that are not normally considered part of their lending practices. Support the solvency of systemically important institutions. The first step is usually to incentivize the private sector to address the problem, often by supporting mergers between failed banks and healthy banks and by regulatory pushes to issue more capital to the private sector. In addition, accounting adjustments can be made to reduce the immediate need for capital to maintain solvency, buying more time for the institutions to earn their way out of their problems. Recapitalize/nationalize/cover losses of systemically important financial institutions. When the above approaches are insufficient to deal with the solvency problems of systemically important financial institutions, governments must step in to recapitalize failed banks. Moving to stabilize
lenders and maintain the credit supply is critical to preventing a crisis from getting worse. Certain institutions are part of the plumbing of the system; one would hate to lose them even if they’re not making money at the moment. It would be like losing a shipping port in a depression because the port goes broke. You want the port to continue to operate and ships to come in, so you have to protect it one way or another—whether through a nationalization, loans, or capital injections.
Debt Defaults/Restructurings Ultimately, the process of cleansing existing bad debts is critical for the future flow of money and credit and for a return to prosperity. The challenge for policy makers is to allow that process to work itself out in an orderly way that ensures economic and social stability. The best-managed cases are those in which policy makers a) swiftly recognize the magnitude of the credit problems; b) don’t save every institution that is expendable, balancing the benefits of allowing broke institutions to fail and be restructured with the risks that such failures can have detrimental effects on other creditworthy lenders and borrowers; c) create or restore robust credit pipes that allow for future borrowing by creditworthy borrowers; and d) ensure acceptable growth and inflation conditions while the bad debts are being worked out. Longer term, the most important decision that policy makers have to make is whether they will change the system to fix the root causes of the debt problems or simply restructure the debts so that the pain is distributed over the population and over time so that the debt does not impose an intolerable burden.
These things rarely happen right away. Policy makers typically fail to recognize the magnitude of the problem initially, instead enacting a number of one-off policies that are insufficient to move the needle. It is only after what is usually a couple of years and a lot of unnecessary economic pain that they finally act
decisively. How quickly and aggressively policy makers respond is among the most important factors in determining the severity and length of the depression. And the question of exactly how these costs are divided between the government (which means the society as a whole) and the bond holders (of varying seniorities), equity holders, depositors, etc., is an important one.
Typically, nonsystemically important institutions are forced to absorb their losses, and if they fail, are allowed to go bankrupt. The resolution of these institutions can take several different forms. In many cases (about 80 percent of the cases we studied), they are merged with healthy institutions. In some other cases, the assets are liquidated or transferred to an “asset-management company” (AMC) set up by the government to be sold piecemeal.
In some cases, policy makers recognize that ensuring the viability of the whole banking system is critical and liquidity and solvency measures are taken at the banking system level. In recent years it has become common for guarantees of bank liabilities to be issued in countries in the developed world. In rare cases, government-financed bank recapitalization is done across all banks, rather than focused solely on systemically important institutions.
There are relatively clear lines for which creditors receive protections:
Small depositors are given preference and experience minimal or no losses (in nearly every case). Often this is explicitly defined as part of a deposit insurance scheme. Coverage is usually expanded during the crisis period in order to ensure liquidity for the banks. Even in cases where there isn’t an explicit deposit insurance scheme, depositors are often prioritized. In around 30 percent of cases studied, depositors did take losses, though they were often on foreign-exchange deposits through conversion at below- market exchange rates. In most cases when institutions fail, equity, subordinated debt, and
large depositors absorb losses regardless of whether the institution was systemically important or not. Protection of senior and subordinated debt holders, and equity recapitalizations that are simply dilutive to existing equity holders, are seen mostly in the developed world. Sometimes policy makers prioritize domestic creditors over foreign creditors, especially when their loans are to private-sector players and are lower down in the capital structure. This is especially true as deposit insurance programs run low on funds. But at the same time governments often end up prioritizing the payment of loans from multinational institutions like the IMF and BIS, as it’s important to maintain availability of support from these public entities, who effectively act as lenders of last resort to countries under stress.
Typically, the process of dealing with the failed lenders is accompanied by a spate of regulatory reforms. Sometimes these changes are modest and sometimes they are very large; sometimes they are for the better and sometimes they are for the worse. They range from changes in how banks operate (e.g., putting in deposit guarantees in the 1930s or Dodd-Frank and the Volcker rule in 2010 in the US) to labor market reforms, from requiring banks to improve credit standards to opening the banking system to competition (including foreign entrants) to raising capital requirements and removing protections to lenders.
Politics plays a big role in determining what reforms are made. In some cases, such reforms end up distorting private-sector market-based incentives on the flow of lending, which can limit the flow of credit to creditworthy borrowers and/or increase the risks of future credit problems emerging. In other cases, they improve the flow of credit, protect households, and reduce the risk of debt problems in the future.
There are two main ways in which failed lenders’ assets or existing lenders’ bad assets are managed: They are either a) transferred to a separate entity (an AMC)
to manage the restructuring and asset disposal (about 40 percent of the cases studied) or b) they remain on the balance sheet of the original lending institution to manage (about 60 percent of cases). And there are several main levers for disposing of the nonperforming loans: a) restructuring (e.g., working out the loans through extended terms), b) debt-for-equity swaps and asset seizures, c) direct sales of the loans or assets to third parties, and d) securitizations.
The use of an AMC generally accelerates the management of the debt problem because it frees up existing banks to return to lending and helps consolidate the bad debts to a centralized entity that can manage sales and restructurings. Selling assets to AMCs also often serves as a mechanism to engage in transfers to banks by pricing them at above market prices. AMCs are typically publicly owned entities that are mandated to sell the assets within some targeted time frame (e.g., 10 years) while minimizing costs to the taxpayer and disruptions of asset markets. They do this by seeking to quickly sell off the performing assets